ECR-R score interpretation: reading your result

The centerpiece of your result is a map built on the two scales of the ECR-R (Fraley et al., 2000). The horizontal axis is avoidance of closeness, and the vertical one is attachment anxiety. The farther right your point, the more distance you tend to need; the higher it is, the stronger your fear that someone close will pull away.

Two lines cross the map at the average scores of a large sample and split it into four areas: secure at the bottom left, anxious at the top left, avoidant at the bottom right, and fearful-avoidant at the top right.

The style name only tells you which area your point fell into. Where it sits inside that area matters just as much. A point right next to where the lines cross and one in the far corner get the same name, yet they describe very different experiences. So first check how far your point is from the lines, and only then read the description. For the exact coordinates, open the collapsed “Your scores in numbers” panel under the map.

Each axis comes with two numbers. The first is your score: the average of your answers to the 18 statements on that axis, on the same 1-to-7 scale you answered on. Attachment anxiety and avoidance scores don’t say much on their own. An avoidance score of 3 looks low because it’s below the middle of the scale, yet it’s almost exactly average: about half of people score lower.

The second number, the percentile, answers a different question: how many people in the sample scored below you. “Higher than 80% of people” means that out of every hundred participants, eighty scored lower. A percentile isn’t a percentage of a style. Some quizzes report that you’re a certain percent anxious or secure, but the ECR-R has no such shares: just two separate scales and where you fall on each.

Next to the percentile you’ll see a level label: below the 30th percentile it reads “low,” above the 70th “high,” and in between “around average.” These labels make the result easier to read; they don’t sort people into different kinds. Two people just above and just below the 70th percentile are barely different.

Your scores are compared with Fraley’s online sample, collected in the early 2000s from at least 17,000 participants. Where those norms come from, and why they call for some caution, is covered in our guide to the ECR-R questionnaire. If you gave your age, a second line compares you with people your age. It usually differs from the first by only a few percentile points, because average scores in this sample change little with age. Your style itself is always based on the whole sample, with no age adjustment.

The label under your style name shows how firmly your point sits in its area. It’s based on the distance to the nearest line, measured in the sample’s standard deviations. A distance under 0.25 of a standard deviation reads “On the border,” 0.25 to 0.6 reads “Mild,” 0.6 to 1.1 “Clear,” and anything beyond that “Pronounced.”

Strength isn’t a grade. A pronounced secure style doesn’t make someone a model partner, and a pronounced anxious style doesn’t spell trouble. It only measures how far your pattern sits from the neighboring ones. With a mild style, you’ll probably recognize traits of the style across the line too, and during a hard stretch your point may cross it.

A borderline result looks like this: a second style appears next to your style name, and on the map your point lies inside the dotted border zone. If both scores are near the lines, your point sits close to the middle of the map, with several neighboring styles at once. Why the lines are drawn where they are, and how often borderline results come up, is explained in the overview of the four styles. When it comes to your own result, something else helps more: read both descriptions and mark the sentences where you recognize yourself.

Sometimes a note about answer quality appears above your result. It isn’t a criticism: the test has simply spotted a pattern that can make the scores less precise. Only the 36 answers from the first part are checked, and your full result is shown either way.

1. 1

   The same answer in a row. 12 or more statements in a row got the same answer. That tends to happen when someone answers on autopilot.

2. 2

   Lots of answers in the middle. Half or more of your answers were a 4, the middle of the scale. That can be honest uncertainty, or it can be fatigue.

3. 3

   Answers that disagree. Avoidance statements with opposite meanings got similar answers, or statements with nearly the same meaning got very different ones. Usually this means some statements were read differently than intended.

4. 4

   Too fast. Less than 2 seconds per statement on average. That’s barely enough time to finish reading one.

If you recognize yourself in the description, you can trust the result while keeping the note in mind. If you rushed or were distracted, take the test again at a calmer moment and compare.

The second part of the test asks about specific people, and a separate map gives each relationship its own point: mother, father, best friend, partner, and close relationships in general. If you skipped a section, you’ll see a note in its place; the other points aren’t affected.

What matters most on this map is what each point is compared with. Not your main result, and not your other relationships, but what’s typical for that kind of relationship (Fraley et al., 2015). And what counts as typical varies: on average, people keep more distance from their father than from anyone else close to them, and the least from their best friend. So the same avoidance score can be “typical” for your father and “higher than typical” for your best friend.

The labels “lower than typical,” “typical,” and “higher than typical” use the same cutoffs as the axis levels, the 30th and 70th percentiles. The partner point is estimated from your answers in the first part, because the test has no separate partner section. Treat it as an approximation. Differences between relationships are normal: attachment is partly specific to each relationship (Fraley et al., 2011). Often the most telling point is the one that stands apart from the rest.

When two tests disagree, the reason is usually how they’re built, not that one of them got you wrong. Here are the most common reasons.

• Picking a ready-made description. In early studies, people chose one of three paragraphs (Hazan & Shaver, 1987), and more than half called themselves secure. Dividing people along two scales gives different shares: on this site, roughly 32% of people land in the secure area.
• A different cutoff. Some tests split each scale right down the middle, at 4. In Fraley’s norms, average avoidance is about 2.92, so an avoidance score of 3.5 lands above our line but below a midpoint cutoff.
• Short forms and untested questions. In a 12-item questionnaire, each answer carries three times the weight it has in the full 36-item version, and questions that were never tested on large samples may be measuring something else.
• A different moment. Answers about closeness depend on what’s happening now: a new romance, a fight, or a breakup can shift them. It also matters which relationship you had in mind while answering.

A fairer comparison looks at your position on the axes, not at style names. Anxiety and avoidance underlie most adult attachment questionnaires (Brennan et al., 1998), so their meaning carries over from one test to another.

If you take the test again a few weeks later, your result will usually look almost the same: over short periods the ECR-R scales are stable, and the figures are on our page about the method. Small shifts are still normal. A change of a few tenths of a point, or of a few percentiles near the middle, can happen even when nothing in your life has changed.

Retaking makes sense after a few months or after a real change: a new relationship, a breakup, a long stretch of therapy. Answer as honestly as you did the first time and keep the same relationship in mind, or the comparison won’t tell you much. Your earlier results stay on the same device, and any you’ve saved to your account will still be there if you switch browsers.

When you compare, look at the direction: has your point moved toward the center of the map or away from it, and has your anxiety gone up or down? What helps that kind of movement, and how long it usually takes, is covered in our article on whether attachment style can change.

Your result describes how close relationships feel to you right now, based on your own answers. It isn’t a diagnosis. An attachment style isn’t an illness or a disorder, and an online test doesn’t replace talking with a professional.

It doesn’t say who’s to blame when a couple struggles, and it doesn’t predict whether a relationship will last. It can’t tell you a partner’s style either: that takes their own answers. And no style makes a person better or worse than anyone else. Every style has its strengths, and your result lists them right next to the difficulties.

It also makes a poor weapon in an argument. “You’re avoidant” rarely helps two people work something out; talking about what you yourself feel usually does more. If a relationship involves put-downs, threats, or any kind of violence, attachment styles are beside the point: your safety comes first.

1. 01Can I show my result to a therapist?

   Yes, and it can be a good way to start a conversation. Send the link to your result or save it as a PDF: it includes your scores, percentiles, and relationship map. Not every therapist works with attachment theory, but the two axes make sense to clinicians from different schools. Just keep in mind that the test reflects how you see your own relationships and doesn’t diagnose anything.
2. 02What if my result upset me?

   Give yourself some time, and don’t treat the result as a judgment of you as a person. A style describes habits that grew out of experience, and habits can change. Reread the strengths of your style. If you’re feeling low or anxious right now, talk with someone you trust or with a mental health professional.
3. 03Where does the partner point come from if there was no partner section?

   From the first part of the test. Almost all of the partner statements in the ECR-RS repeat ECR-R statements, so we take your answers to those and compare them with the norms for partners. It’s an approximation. If you answered about a past relationship, the point describes that one.
4. 04Why doesn’t my “close relationships in general” point match my main style?

   Because they answer different questions. Your main style comes from statements about romantic relationships, while the general section asks about all the people who matter to you at once and is compared with its own norms. A gap usually means love feels different to you than closeness in general, and that’s common.